The recent crisis exposed a fundamental flaw with our banking system (actually it’s been there since 1934, but this was the worst manifestation of the problem.) Because of various government backstops (FDIC, F&F, TBTF, etc) our financial system takes excessive risks.
Tim Congdon has an intriguing proposal to address this problem. His plan would not eliminate moral hazard, but it just might reduce it:
Suppose that banks are short of capital because of bad asset selection and heavy losses, and that there is a risk of deposits not being covered by banks’ remaining assets-the situation allegedly facing Britain’s banks and economy in late 2008. Surely everyone ought to agree that, ideally, the job of replenishing banks’ capital should fall on the private sector, not the state, and that as far as possible accountability must lie in only one place. How might that be achieved?
I propose a structure which is almost the polar opposite of Buiter’s. In a 2009 monograph, Central Banking in a Free Society, I argued that the capital of the Bank of England should be provided by the leading banks, and that the functions of the central bank and deposit insurance agency should be amalgamated. Whereas Buiter wants, in effect, to nationalise the arrangements (and the costs) of cleaning up a banking system fiasco, I would privatise them.
My proposal is much less radical than it may seem. Recall that at present the banking industry as a whole can in fact be levied by the FSCS if its funds are exhausted as a result of losses at one or a handful of banks. So, making the commercial banks the shareholders in the central bank and giving the central bank an explicit responsibility to protect deposits, has two consequences. First, the central bank would extend loans only if confident they would be repaid. The well-behaved, risk-avoiding and profitable banks have a strong interest in preventing their risk-prone competitors from incurring losses and ruining the system. Secondly, if the system nevertheless ran into trouble, the first line of defence would lie in the private sector, via the capital-levying power of the Bank of England. If the government did have to come in, it would be only after the bankers had decided that they could not help themselves. It would indeed be a last resort.
I’d like to separate two questions: Should there be a unified institution to handle deposit insurance and emergency capital provision, which is owned and financed by the banking sector? And should this institution be the central bank? I agree with Congdon on the first point, but don’t see the second as being essential, or even politically feasible (at least in the US.) In America, banks already have partial ownership of the Fed (albeit less than people imagine) and that’s highly controversial. So let’s look at the first question.
The beauty of Congdon’s idea is that decisions on bank rescue will be made by those with the right set of incentives. Suppose a bank gets into trouble, but is not a systemic threat. The socially optimal solution is bankruptcy, that’s how markets should work. That sends a signal to the banks that they should take fewer risks. It sends a signal to creditors that they should carefully monitor bank behavior. It would be nice to send a signal to depositors, but that doesn’t seem politically feasible—at best we might be able to trim the maximum coverage a bit.
But what if the bank failure was a systemic threat? In that case the institution would have an incentive to provide emergency capital injections–in order to prevent contagion that would threaten the rest of the banking system. This institution basically internalizes externalities. The money belongs to the banks, who would be shareholders (in some proportion to bank size.) If funds were paid out, banks would have to chip in to re-capitalize the institution.
In once in 100 year disasters the Treasury might have to backstop the institution with Federal loans. But as we saw with the TARP loans, the Treasury could simply require that these loans be repaid by the institution, once the banks got back on their feet.
In theory, taxpayers should monitor Washington regulators, so that they wouldn’t show favoritism to politically important special interests. But we know from public choice theory that that doesn’t work very well. The banks themselves are the best monitor, or perhaps a Board of Directors that is directly answerable to the banks.
Here’s another way of looking at the plan. It’s politically difficult to get the big banks to hold enough capital to overcome the TBTF problem. They’ll whine about picking up and leaving New York, London, or Zurich. But it’s also unlikely that all big banks will get into trouble at exactly the same time. This institution would pool capital in a way the could nip a potential financial panic in the bud, without requiring taxpayer money.
Of course the cost of this insurance fund will ultimately be borne by customers, so in that sense it is a tax. But it has two advantages over a bailout using income taxes. The tax reflects the external costs imposed by fractional reserve banking. And second, the banks would have much more incentive than government bureaucrats to use the funds wisely.
OK, I’m not a banking expert—tell me what’s wrong with this plan?